284 THE APODID^ PART ii 



nida with the Xiphosuridae and Eurypteridae taking 

 the two latter out of the division of the Crustacea. 



If there is any truth in our general argument as to 

 the derivation of the primitive Crustacea from a bent 

 Annelid, and of the Tracheata from an Annelid not 

 so bent, there is no need for any such alteration 

 in the formerly accepted classification. The resem- 

 blances in inner and outer organisation between 

 the Xiphosuridse and the Scorpionidae, striking as 

 they undoubtedly are, we believe to be simply 

 due to the fact that they are both descended from 

 Annelids. The agreement in the number of segments 

 and cephalothoracic limbs is by far the most important 

 argument in favour of the new classification. 



But now it seems to us that it is by no means im- 

 probable that two groups of animals descended from 

 many-segmented Annelids should possess the same 

 number of segments, especially when we find that 

 somewhere about the same number of segments seems 

 to have best suited many other groups belonging to 

 both divisions. The Malacostraca have twenty, the 

 free-swimming Copepoda about fifteen, the Hexapoda 

 sixteen, and many genera of the Myriapoda from 

 fifteen to thirty. 



The resemblance between the limbs of Limulus and 

 Scorpio does not seem to us so great as it is often 

 assumed to be. The five pairs of jaws ranged round the 

 ventral mouth of Limulus, whether our theory of their 

 origin from Annelidan parapodia is true or not, form 

 a feature which has no counterpart in the limbs of 

 Scorpio. This is, to our mind, a most important 



